mcp-vcd

MCP.Pizza Chef: SeanMcLoughlin

mcp-vcd is a Model Context Protocol (MCP) server designed to handle Value Change Dump (VCD) waveform files commonly used in hardware simulation and verification. It allows models to access detailed signal change data from large VCD files without loading the entire file into context, supporting queries by signal name and optional timestamp ranges. This server facilitates efficient, scalable waveform analysis by feeding structured signal change information into LLMs, enhancing debugging, verification, and hardware design workflows. Installation is straightforward via pip, and it integrates seamlessly with MCP-compatible hosts like Claude Desktop.

Use This MCP server To

Extract signal changes from large VCD waveform files Query specific signal data within time ranges Feed hardware simulation data into LLMs for analysis Assist in debugging digital circuit designs Enable scalable waveform data access in AI workflows

README

mcp-vcd

A model context protocol (MCP) server for value change dump (VCD) waveforms.

mcp-vcd-demo.mov

Tools

get-signal: Provide all changes of the specified signal name to the model's context. This is useful for large waveform files with many signals where you cannot fit the entire VCD file into the model's context window.
- Can optionally specify a start and end timestamp range for the model to look at.

Installation

uv pip install --system mcp-vcd

And add the following to your claude_desktop_config.json:

"mcpServers": {
  "mcp-vcd": {
    "command": "uv",
    "args": [
      "run",
      "mcp-vcd"
    ]
  }
}

See Anthropic's MCP documentation for more info.

mcp-vcd FAQ

How do I install mcp-vcd?

Install mcp-vcd using pip with the command `uv pip install --system mcp-vcd` and configure your MCP host to run it as a server.

Can mcp-vcd handle large VCD files efficiently?

Yes, mcp-vcd supports querying specific signals and time ranges, avoiding the need to load entire large VCD files into the model context.

What commands does mcp-vcd provide?

It provides the `get-signal` command to retrieve all changes of a specified signal, optionally within a start and end timestamp range.

How do I integrate mcp-vcd with my MCP host?

Add the mcp-vcd server configuration to your MCP host's config file, specifying the command and arguments to run the server.

Is mcp-vcd compatible with multiple LLM providers?

Yes, mcp-vcd works with MCP clients that support models from providers like OpenAI, Anthropic Claude, and Google Gemini.

What is the benefit of using mcp-vcd over loading VCD files directly?

mcp-vcd enables selective, efficient access to signal changes, reducing context window usage and improving performance in AI-assisted hardware analysis.

Does mcp-vcd support real-time waveform analysis?

While primarily designed for post-simulation VCD files, mcp-vcd can be used in workflows that require incremental or range-based signal queries.

Where can I find documentation for mcp-vcd?

Documentation is available on its GitHub repository and through the Model Context Protocol official site.